Metastatic breast cancer is the leading cause of breast cancer-related deaths. Predictive signatures for poor prognosis breast cancer are available and routinely employed in the clinical setting; however, these signatures describe the risk of metastasis, not the underlying biologic drivers of metastasis. Moreover, accompanying therapeutic approaches to prevent metastasis are not provided by these established signatures. The underlying biology driving metastasis is largely unknown. Identifying genetic drivers of metastatic breast cancer and the timing during which these changes occur is critical to the development of effective therapeutics to prevent and treat metastases; such a study has yet to be performed in human tissues. We hypothesize that specific genetic and/or genomic factors are specifically enriched in breast cancer metastasis, and that some metastatic features are established within primary tumors. Utilizing a Rapid Autopsy Protocol established at UNC Chapel Hill, we have collected 169 metastatic breast cancer tissues from 34 individuals and performed RNA and DNA whole exome sequencing. Aim 1 will define those genetic features enriched in metastases through a statistical comparison of DNA copy number, point mutations, and RNA gene expression changes in metastatic breast cancer tissues as compared to 1000 primary breast cancers previously published through the Cancer Genome Atlas. Aim 1C will then compare DNA copy number alterations and point mutations for each individual patient's collection of metastases to establish a map of the acquisition of molecular changes across time. Patients with breast cancer brain metastases have extremely limited survival and no approved systemic therapies. Current literature from in vivo modeling suggests that significant changes to both breast cancer cells and the tumor microenvironment are critical for metastatic seeding and growth. Aim 2 will establish specific genetic features of breast cancer brain metastasis and the altered signaling in the supporting adjacent brain parenchyma. Comparing gene expression in adjacent non-malignant brain parenchyma to matched distant brain parenchyma both in human tissues and in two in vivo breast cancer brain metastasis models will identify supporting microenvironment signals unique to the adjacent brain parenchyma. Our goal is to identify genetic features enriched in breast cancer metastasis, ultimately to define future therapeutic targets and help decrease mortality from breast cancer. Along with the superb environment at UNC Chapel Hill, exceptional mentoring by Dr. Perou and Dr. Anders, and a comprehensive training plan, this fellowship will provide a critical foundation to build my future career as an independently funded physician-scientist in the field of cancer metastases.